Ethernet cables

Noogle posted:

Remember the one true creed:  bits are bits.  

i wish.

its becoming increasingly clear to me that 'bits are bits' doesnt paint the complete picture. apparently even in this domain, we cannot escape the usual analogue evils -- RF etc...

there are folks here who will explain that much better than i can.

enjoy...

ken

Heresy!  Cue Simon-in-Suffolk...

Just to put some petrol on the fire, that is already burning... I DO hear a difference between streaming cables. Unfortunately bits are no longer bits here...

Guys, I don't want to sound like a worn out record, but this is nought to do with the 'bits' of our sampled audio. This is a  meaningless concept and abstract to the analogue voltages that are being modulated at radio frequencies on the cable. This is an analogue world. Modulating RF current causes EM interactions and line driver PSU modulation no matter how hard we try to mitigate it. These EM interactions will vary with different cable designs as they couple to connected equipment. It is RFI in its varying capacities and its interactions we hear... and therefore each cable design WILL ALMOST  'sound' different to varying degrees ... But it is nothing to do with the carried digital payload.

So Ken you are almost right to my mind.. But I would say in this area, the bits don't paint any of the picture....

Simon-in-Suffolk posted:

Dan43 posted:

Melco ethernet cables :  http://melco-audio.com/products.html#c1ae

Any thoughts on these, £60 for 0.5m?

That sounds ridiculous.. High quality low noise cable seemingly aimed at the multimedia industry and for constructing Cat 6 patch leads / infrastructure wiring is about £1.50 per metre from industrial cable stockists.

i thought one of the advantages of the Melco was you didn't need to use boutique cables, as the common mode noise had been removed by the Melco filter?

Which is what I was thinking also, the Melco N1A is doing all this anyway.

UHES supplied my NDS system at the time and installed their own ethernet cable (I may conclude off the shelf, but maybe they bought in from a supplier a particular brand of cable but I wouldn't know) Cat6e cable with the same terminated ends as the Melco cable, from their own cable drum and crimping the ends in situ, and supplied it for free as part of the system installation (2 x 3m lengths) perhaps verifying your thoughts on the cost Simon. Needless to say these cables have provided me no reason to change that I can fathom.

Mike-B posted:

" ...........   has wideband transmission speed 6 times that of Cat 5e, 2.4 times of Cat 6."

So I wonder what happens to all that "speed" when it crashes up against the 100MHz port on the Naim box

 

Mike.. Don't confuse 100 MHz with 100 Mbps 

The actual line frequency of 100 BaseT is effectively approx 31 MHz

So the only conclusion I can see with that high bandwidth that is that any waveform filtering within the required bandwidth will be minimal.

Yes Simon,  we've been there before,   I do try to keep it simple  Melco are taking 600MHz,  so maybe best not go to mbps to add more confusion.  

Re your EM interaction  ........ One or two cable supplies talk about near end crosstalk, return (reflection) loss & skew delay that potentially can cause data transfer issues.  I guess this is much the same as your EM interaction (??)

Mike .. To may way of thinking ..  yes

Let's not confuse the medium with the message.  I can write the binary number 1010 on vellum using a £1000 Mont Blanc fountain pen with a gold nib, or write it on the back of a fag packet using a 10p Bic biro.  As long as you can read it, It's still the same number.

Simon-in-Suffolk posted:

...

So Ken you are almost right to my mind.. But I would say in this area, the bits don't paint any of the picture....

A picture analogy...

Ethernet distribution of media is like a jigsaw, the whole picture is delivered in little chunks.  However, unlike a jigsaw, all the pieces are numbered on the back, so you know where they go.  The are also normally delivered in sequence, but if one goes missing then you know because the numbers are out of sequence, and you can ask for the missing piece again.  When you put the pieces together in the buffer and read it in sequence, the media signal (i.e. the music) is reconstructed.

The bits don't paint the picture until they are read out of the buffer, at the end and in the right sequence; while they are being transmitted across the network, they don't have to make any sense at all (at least not to a human!).

Simon-in-Suffolk posted:

i thought one of the advantages of the Melco was you didn't need to use boutique cables, as the common mode noise had been removed by the Melco filter?

Ditto if you use a usb isolator between renderer and DAC (or isolator/SPDIF convertor ets)

Innocent Bystander posted:

Simon-in-Suffolk posted:

i thought one of the advantages of the Melco was you didn't need to use boutique cables, as the common mode noise had been removed by the Melco filter?

Ditto if you use a usb isolator between renderer and DAC (or isolator/SPDIF convertor ets)

Lord been looking at ethernet isolators the WWW indicates Medical Isolators a being the best adhering to IEC 60601-1 standard for all manner of quality reasons. 

From Wiki : Network connections between medical devices and Ethernet networks (and other non-medical devices, such as personal computers) must be in accordance with the IEC standard IEC 60601-1 (3rd edition)

What a subject this has become, but good to dig into things a little.

To break ground loops, I use a CAT5e unshielded coupler.  Simple cheap effective.

I don't need total electrical isolation (including the twisted pairs) as I don't have large electrostatic potential differences between different parts of my network; so I let the high CMRR of Ethernet interfaces deal with the signal.

IEC 60601-1 is all about product safety specifically for medical equipment.  Nothing much to do with small home networks.

Mike,

Rather than your soggy moggies, I've decided to standardise on ferrets and ferrites:  The ferrites are needed to filter the signals as ferrets are kleptomaniacs and aren't very selective in what they carry around!  I don't think ferrets are particularly welcome in medical scenarios, so I feel IEC 60601-1 isolators are overkill.

I used to know a ferret and moggie that would play together.  The moggie would sit at the top of the stairs and the ferret would climb up the stairs.  When it got to the top step the moggie would give the ferret a well timed thump (claws in), causing it to fall back down the stairs like a rather out of control slinky!  After it picked itself up at the bottom, it would climb the stairs again, with the same result!

Noogle posted:

Let's not confuse the medium with the message.  I can write the binary number 1010 on vellum using a £1000 Mont Blanc fountain pen with a gold nib, or write it on the back of a fag packet using a 10p Bic biro.  As long as you can read it, It's still the same number.

I like this analogy - if the ink fades in the sunlight then you can't reliably read what was written.. and if the paper is fibrey then what was written becomes less clear as the ink get drawn in the paper fibres

Huge posted:

I used to know a ferret and moggie that would play together.  The moggie would sit at the top of the stairs and the ferret would climb up the stairs.  When it got to the top step the moggie would give the ferret a well timed thump (claws in), causing it to fall back down the stairs like a rather out of control slinky!  After it picked itself up at the bottom, it would climb the stairs again, with the same result!

What a great story. I once had a friend who kept a buzzard in his bedroom (I know, it's weird) and one day a visitor made the mistake of entering the house wearing a fur hat. The buzzard spied the hat, flew down the stairs, grabbed the hat and ripped it to pieces. The only connection to the thread is that this is a good 'animals and stairs' story. My nephews kept ferrets when they were younger though, if that helps the ferretreferenceometer score. 

Simon-in-Suffolk posted:

Noogle posted:

Let's not confuse the medium with the message.  I can write the binary number 1010 on vellum using a £1000 Mont Blanc fountain pen with a gold nib, or write it on the back of a fag packet using a 10p Bic biro.  As long as you can read it, It's still the same number.

I like this analogy - if the ink fades in the sunlight then you can't reliably read what was written.. and if the paper is fibrey then what was written becomes less clear as the ink get drawn in the paper fibres

And if someone scribbles over it the writing becomes less clear...

But in both cases, unless too far gone, there are usually ways of getting to the original message and accurately reading the number.